Sign in to follow this  
mcompengr

A.I.'s dirty little elephant in the room.

Recommended Posts

Life as a mixture of hardware and software is a fine model, but heat-sinks, power supplies, I/O and
special relativity notwithstanding, any (binary) digital computing system, any computation can be
implemented with just hardware, 100%, and that with just the meek and lowly nand gate (and-not).
[(Speculation: the reverse may be almost true, but for implementation purposes some kind of "hardware"
would seem to be needed.)] In any case bit-slicing, microcoding, FPGA, or just any memory used to
implement a function, etc., these things have done away with any line between hardware and software.

Good luck with "free will". If nothing else, boolean logic is deterministic.

Share this post


Link to post
Share on other sites

An external source of randomness(and necessarily requiring hardware) is the answer imo.

 

...and why is the NOR gate so unloved?

Share this post


Link to post
Share on other sites

I forget exactly how close, but there are some functions that

can not be implemented using only NOR gates.

 

-Martin

Share this post


Link to post
Share on other sites

 

I forget exactly how close, but there are some functions that

can not be implemented using only NOR gates.

 

Since you can make any nand gate from nor gates this can't be true unless it is also true of nand gates.

 

In fact both nand gates and nor gates are 'universal' gates.

 

http://en.wikipedia.org/wiki/NOR_logic

Share this post


Link to post
Share on other sites

So, why didn't you tell us sooner? No, really thanks!!

 

Ok, any thoughts on external randomness? I thought

generally of that, and what if the randomness came

from human input? A super-mind, all yours and at your

finger tips??

 

So, randomness from gates? No two gates trip at the

same speed and might vary somewhat. So combinational

logic might. Sequential is out, I suppose. Or is it the other

way around?

 

This is fun!

 

-Martin

Share this post


Link to post
Share on other sites

Perhaps this whole business comes from a simple fact.

All our present-day computers, are made of a few simple elements. Mainly copper and silicon. They may have some additional elements, but they're basically made of metal.

 

Now if that's true, how could these contrivances, possibly give rise to to "intelligence"? Wouldn't it be like expecting an alarm-clock to understand the meaning of time. Or a CD player to have an appreciation of music and speech. The CD player certainly reproduces music and speech. In a limited way. But it doesn't know what it's doing. It doesn't know anything - it's just a metal mechanical contrivance.

 

As are modern computers - they're just metal mechanisms. They work by sending electrons through NOR and NAND gates, following their program. Just like a light-bulb lights up, or goes dark, according to which way its switch gets pressed. And the switch itself, is a simple "on-off" mechanical device, made mostly of metal. This shows that metal atoms haven't got enough complexity to become intelligent.

 

So it's a false trail to expect AI from metal computers. It'd be like making a bronze replica statue. Then expecting it to come to life, and behave intelligently. It can't - because it's not complex enough.

 

So, shouldn't AI researchers ditch metal. And concentrate on building carbon-based devices? The carbon atom is chemically vastly superior in its ability to form complex molecules. These could possibly result in an Artificial Carbon Intelligence.

Edited by Dekan

Share this post


Link to post
Share on other sites

From computability theory, we know that anything that can be computed by any (Turing equivalent) computer can be computed by any other. So switching from silicon to carbon (or from electronics to brains) doesn't change what a computer is able to do.

Share this post


Link to post
Share on other sites

CD player, or alarm-clock, are doing their very limited program (which might not be program but just set of automatically executed tasks).

While real AI computer program can/should learn.

Observe and analyze world by sensors and react/adopt to them, remembering how world behaves.

Edited by Sensei

Share this post


Link to post
Share on other sites

Would your analysis be the same from the equally accurate, but more precise

(as to potential success) perspective of conscousness rather than intelligence?

 

-Martin

Edited by mcompengr

Share this post


Link to post
Share on other sites

So, why didn't you tell us sooner? No, really thanks!!

 

Ok, any thoughts on external randomness? I thought

generally of that, and what if the randomness came

from human input? A super-mind, all yours and at your

finger tips??

 

So, randomness from gates? No two gates trip at the

same speed and might vary somewhat. So combinational

logic might. Sequential is out, I suppose. Or is it the other

way around?

 

This is fun!

 

-Martin

Radioactive decay comes to mind. There has actually been a die rolling system created based on a similar principle (they used photons in some manner i believe)

Cosmic background radiation may be another good option. While gates trip at different speeds, I imagine there would be some bias, in that a gate that trips quicker than another will tend to continue tripping before the other, making for a rather predictable circuit. Although there are a bunch of variables that might play into it, heat being the one that comes to my mind. Perhaps a mess of transistors whose operation changes its own heat distribution, thereby changing its behaviour?

Edited by DJ027X

Share this post


Link to post
Share on other sites
On ‎1‎/‎2‎/‎2015 at 11:53 PM, DJ027X said:

Radioactive decay comes to mind. There has actually been a die rolling system created based on a similar principle (they used photons in some manner i believe)

Cosmic background radiation may be another good option. While gates trip at different speeds, I imagine there would be some bias, in that a gate that trips quicker than another will tend to continue tripping before the other, making for a rather predictable circuit. Although there are a bunch of variables that might play into it, heat being the one that comes to my mind. Perhaps a mess of transistors whose operation changes its own heat distribution, thereby changing its behaviour?

And now, three years later quantum computing is on steroids.

We MUST throw money at it because "they" might try to build one.  My father worked on nuclear aircraft.  Somebody had to do it.  A.I. is making money and it's fun, and scientists will follow their bliss anyway.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Sign in to follow this